Characterizing Forest Edge Structure Globally
Abstract
More than 70% of the world's forests are within 1 km of a forest edge. Forest edges experience increased sunlight and warmer and drier microclimates. They are frequently exposed to human activity and are prone to fires and storm damage. There are only a few studies of forest structure at edges, the majority at the plot scale, with some disagreement about the magnitude and direction of change in forest structure from interior forest to the edge. This poor understanding of forest edges leads to inaccuracies in forest modeling, conservation planning and carbon monitoring. The goal of our research is to improve the global characterization of forest edges and their impacts on biodiversity.
To conduct our research, we quantified forest structure using a 30 m resolution global fractional tree cover product derived from 2015 Landsat data. We classified each pixel as forest or non-forest, then calculated more than 16 landscape variables globally (e.g., distance to edge, patch size, average forest area per 10 km2). We then analyzed the relationship between landscape variables and fractional forest cover. Our results show that forest density decreases at forest edges across ecosystems, though local variability is high. Forest edge effects are frequently significant beyond 1 km into the forest. Comparing forest density with adjacent land cover shows that forest edges near croplands are similar to those near developed areas, both having a large decrease in forest cover near the edge. Natural edges are more variable and may increase in forest density near water bodies. We give an example of the disproportionate loss of interior forest caused by many small patches of deforestation throughout West Virginia forests due to clearing for oil and gas extraction. We expect our work to contribute to improved forest and carbon modeling and to more effective conservation efforts, particularly in preventing the loss of threatened interior forest species.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2020
- Bibcode:
- 2020AGUFMB060.0001D
- Keywords:
-
- 0410 Biodiversity;
- BIOGEOSCIENCES;
- 0466 Modeling;
- BIOGEOSCIENCES;
- 0480 Remote sensing;
- BIOGEOSCIENCES;
- 1922 Forecasting;
- INFORMATICS